James Larus, leader of Microsoft's Singularity OS project, departs company

The man who lead the development of an operating system project at Microsoft other than Windows has decided to leave the company after 16 years. James Larus, who was a principle researcher at Microsoft Research, will become the new dean of computer science at Ecole Polytechnique Federale De Lausanne in Switzerland starting in October.

ZDNet reports that Larus, who joined Microsoft in 1988, was the leader behind the Singularity OS project, which was based on managed code and was created from scratch. The OS started as a research project in 2003 but was never used in a commercial product. Microsoft stopped development on the OS in 2008 but decided to make the work on the Singularity OS available for download for academic, non-commercial projects. Larus later helped to lead Microsoft's eXtreme Computing Group, which developed cloud-based hardware and software products.

Although the Singularity OS is no more, ZDNet.com reports that many other operating systems at Microsoft were directly or indirectly influenced by Singularity. One of them is Midori, which is still in development at Microsoft, even though the company won't officially talk about it. Midori could still be used inside some kind of Microsoft commercial product, but at the moment there's no word on when that might happen.

Source: ZDNet.com | Image via Microsoft

Report a problem with article
Previous Story

Microsoft Research helping to create new short looping videos

Next Story

Rumor: Ashton Kutcher may soon be promoting Lenovo Windows 8 PCs

16 Comments

Commenting is disabled on this article.

Spicoli said,
Dean of a CS department is going to be a big culture shock from a commercial business. Way more politics and way less productive work.

But in an academic environment you can pursue pure researches something that in a commercial environment would be impossible.
What such kind of researches generate are new theories and ideas which, although not necessarily in the short terms, are what have always fueled human progress.

Fritzly said,

But in an academic environment you can pursue pure researches something that in a commercial environment would be impossible.
What such kind of researches generate are new theories and ideas which, although not necessarily in the short terms, are what have always fueled human progress.


"pure research" in a commercial environment is not impossible. Look at most of the work MSR does.

SharpGreen said,

"pure research" in a commercial environment is not impossible. Look at most of the work MSR does.

Yes MSR indeed pursues many "abstract " researches but it is an more an exception than the norm.

Spicoli said,
Dean of a CS department is going to be a big culture shock from a commercial business. Way more politics and way less productive work.

Not necessarily - many universities are tied in with businesses that work together on projects such as long term R&D. My brother, as part of his doctorate in material science, worked with a business specialising in methanol looking at developing new metal pipes that can last longer at the production facilities and to also test existing designs for flaws.

Fritzly said,

Yes MSR indeed pursues many "abstract " researches but it is an more an exception than the norm.

Sadly, I would argue the opposite is true. Most of the work at MS Research has no product expectations and often is freely released to the technology world.

Microsoft purposely gives up opportunities from MS Research projects by either giving them away or shelving them for several years due to heavy lifting it would take to get them to market as a product in the current state of technology.

Often the competition it would create with partner technologies would be detrimental and it is better to offer the technologies to the partners or just release a paper on the technologies.

Unless they are working on a specific production product, like the MS hardware engineers working on the Xbox team, the majority of work is assumed to be 'abstract'.

Mobius Enigma said,

....

I believe MSR is a way for MS to give back to technology. Help it progress.

Most (if not all) giant MS haters, have MS developed technology in their hardware or software.

Worst, or best, is that a lot of things MSR was working on in the 90s, have come true in the last decade.
And the things I sometimes come across now keeps me very intrigued.

Singularity was always a small research project at MS, it was a test to see how good a fully managed OS could be, I have no doubt that lots of the managed API ideas that they came up with in the project found their way into what we know of as WinRT and Metro right now.

It was never going to turn into a commercial OS though, they'd have to fully ditch Win32 in that transition and lots of other low level changes have to happen to, drivers and so on from what I remember.

Singularity is believed to be the base for the development of Midori.
And Midori is and has been used to improve NT. Several new under the hood changes in 7 and 8 seem to come from Midori.

So even though it was never meant to be turned into a commercial OS(more so, its open source). It has helped develop their flagship OS.

AWilliams87 said,
Word is it's Midori which is set to supersede Windows.

Concepts from Midori are already present in Windows 8. Additionally, work from Midori is used in Windows NT Azure.

NT is far more extensible than people realize, and most of these OS design concepts can be integrated into NT. It is like the video subsystem overall that came out of the Xbox team, NT was able to seamlessly integrate the WDDM even with key features needing to be a part of the core NT scheduling in the kernel. (With Windows Vista/7 the WDDM and the older XPDM existed both at the same time, with the kernel able to flip between and use the different models natively equal.)

With all the OS projects at Microsoft, it will be several years before NT has to be replaced; instead, NT will slowly continue to evolve using the newer features that come from these projects.

('Inside NT First/Second Edition' is a great place to start and see why NT can adapt in ways traditional kernels and OS models cannot.)

More like he choose a peaceful academic life in Switzerland and who can blame him after 15 years in the business world, especially as he is an academic and researcher.